An Intelligent Deep Learning Framework for Early Breast Cancer Detection Using Medical Imaging
IOSR Journal of Electronics and Communication Engineering · 2026-05 · 21 (5) · 09-20
DOI: 10.9790/2834-2105010920 ↗Kurz gefasst
Fachartikel in der Zeitschrift IOSR Journal of Electronics and Communication Engineering, 2026-05. Verfasst von Abhinav Pratap Singh und 2 weiteren Autor:innen.
Abstract (Originalsprache)
The early and correct diagnosis of breast cancer is important for better clinical results and prompt treatment planning. The analysis of medical images, however, is complex due to complex tissue structures, variations in image quality and subtle malignant patterns. In this paper, an intelligent deep learning framework is proposed for automated breast cancer detection from medical imaging, called ViT-GWO. The proposed framework is a combination of a Vision Transformer (ViT), which is used to learn global contextual and discriminative representations from medical images of the breast, and Grey Wolf Optimizer (GWO), which is used for selecting the most relevant subset of features from the high dimensional feature space. The methodology comprises of image preprocessing, image normalization, low-pass filtering, data augmentation, image patch generation, patch embedding, positional encoding, Transformer-based feature extraction, and GWO-based feature optimization. The optimized features are then given to a SoftMax classification layer which is used to separate benign and malignant breast tissue. It is tested on breast cancer data sets DDSM and MIAS, and the results are compared with the results of the CNN, MLP, and EML methods. The experimental results show that the proposed ViT-GWO framework model has an accuracy of 99.57%, a precision of 99.25%, a recall of 98.76%, MCC of 99.89%, and F1-score of 99.77%. Based on these results, it can be concluded that the use of Vision Transformer network for generating representations and GWO for optimizing those features can be a promising approach for automated breast cancer detection and classification. The proposed framework has the potential to enhance classification accuracy in computer-aided diagnosis, whilst minimizing the effect of redundant features.
Zitieren
APA
Singh, A. P., Pawar, D. R., & Mohan, D. A. (2026). An Intelligent Deep Learning Framework for Early Breast Cancer Detection Using Medical Imaging. IOSR Journal of Electronics and Communication Engineering, 21(5), 09-20. https://doi.org/10.9790/2834-2105010920
Vancouver
Singh AP, Pawar DR, Mohan DA. An Intelligent Deep Learning Framework for Early Breast Cancer Detection Using Medical Imaging. IOSR Journal of Electronics and Communication Engineering. 2026;21(5):09-20. doi:10.9790/2834-2105010920
BibTeX
@article{Singh2026,
author = {Singh, Abhinav Pratap and Pawar, Dr. Reeta and Mohan, Dr. Ashwala},
title = {An Intelligent Deep Learning Framework for Early Breast Cancer Detection Using Medical Imaging},
journal = {IOSR Journal of Electronics and Communication Engineering},
year = {2026},
volume = {21},
number = {5},
pages = {09--20},
doi = {10.9790/2834-2105010920}
}
EUVIMED